Microtalk
Why Real-Time Systems Need a Real-Time Database
Steven Graves
09:58
Watch this session
This EOC 2022 session is in the archives. It is included in the EOC 2026 + Archives pass, with all 504 sessions since 2020. Passes bought now also include the upcoming EOC 2027 + archives until December 31, 2027.
Get your pass · $395
Already registered? Log in
Historically, real-time systems and database systems have been like oil and water. The reason? Making an API call to a non-deterministic database system could cause a real-time task to exceed its deadline. Previously, real-time systems were not overly complex and could do without database support. Today, however, real-time systems’ tasks need to collect, aggregate, correlate and analyze data from disparate sources (sensor data fusion) and could benefit greatly from a shared repository. This MicroTalk will present a solution: adding time-cognizance to database transactions, and suitable transaction schedulers.
1 / 3
Please log in or create an account to test your knowledge and see the answers.
What is the main reason Steven Graves argues that real-time systems need a real-time database (rather than a conventional high-speed database)?
A
Because real-time databases provide deterministic, time-cognizant responses that can guarantee meeting transaction deadlines.
B
Because real-time databases always provide much higher raw throughput than conventional databases.
C
Because real-time databases eliminate the need for sensor data fusion by pre-processing sensor inputs at the database layer.
D
Because real-time databases reduce storage costs by aggressively compressing time-series data.
E
Because real-time databases remove the need for a real-time operating system and its scheduling guarantees.
No comments or questions yet. Be the first to start the conversation!